How do you design a canteen where people eat well without being told to?
Google noticed its employees were eating too much of the free office candy, and rather than launching a campaign about it, the company ran an experiment. They kept the chocolates, but moved them into opaque containers, out of plain sight. At the same time they put dried figs, pistachios and other healthier snacks into clear glass jars, prominently displayed. Nothing was banned. Nothing was lectured. Only the visibility changed.
In the New York office alone, employees consumed 3.1 million fewer calories from M&Ms over seven weeks. No campaign. No promotion. Just placement.
The canteen that argues with itself
The traditional approach to a workplace canteen is almost designed to fail. Add a healthy range. Put the nutritional information on the menu card. Offer a vitality programme as compensation for whatever the canteen is doing to people the rest of the week.
The trouble is what the room is actually saying while all this goes on. The healthy options sit at the back. The fat and the salt sit at the front. The pricing rewards the wrong choice. And then the same organisation pays for a wellness programme to repair the damage it engineered into the layout. The canteen argues with the wellness budget, and the layout wins, because the layout is present at the exact moment of choice and the wellness programme is an abstraction filed away for some healthier future self.
Google's move sidestepped all of it. They did not add a healthy range and hope. They did not moralise about the M&Ms. They asked what the room was making easy, and then changed what the room made easy. The candy did not disappear; it simply stopped being the path of least resistance. The healthy snacks did not come with a sermon; they just became the visible, obvious, reach-for-it option. The choice architecture did the persuading that no poster could.
Why this is design, not instruction
This is the cleanest possible illustration of the line between design and motivation, which is exactly why it is worth being careful about.
Google did not motivate anyone. Nobody was told to eat fewer M&Ms. There was no health campaign, no challenge, no nudge in the hectoring sense of the word. The employees were not asked to want anything different. If anything, the experiment worked precisely because it bypassed wanting altogether. People still wanted the chocolate. They just reached for it less, because it was no longer sitting in a clear jar at eye level inviting the automatic hand.
Motivation would have tried to change what people wanted. Design changed what was easy to reach. And reaching, it turns out, is mostly automatic. We do not deliberate over a handful of M&Ms; we grab them because they are there, visible, frictionless. Make them slightly less visible and slightly less frictionless, make the figs the easy grab instead, and behaviour shifts without anyone deciding to be healthier. The 3.1 million calories were not a triumph of willpower. They were a triumph of placement.
The eating behaviour was never really about the discipline of Google's staff. It was about where the jars were. At SUE, the Influence Framework and SWAC are the methods we teach in the Behavioural Design Fundamentals Course to systematically analyse how placement and salience shape behaviour before any conscious decision is made.
Amsterdam
Rated 9.3 out of 10. Certificate included.
The principle: choice architecture and salience
The idea underneath this has a name, and naming it is what turns a clever office tweak into something you can apply anywhere people choose.[1]
Choice architecture is the term, defined as altering the placement or properties of options within an environment in order to shift behaviour, without removing any options or changing what is on offer. Its close partner here is salience: how visible, how prominent, how immediately present an option is. The core finding across this body of research is consistent and slightly humbling. The position, visibility and accessibility of an option shape what people choose more powerfully than their conscious intentions do.
That is the part worth sitting with. We like to believe our choices flow from our values, our health goals, our considered preferences. Sometimes they do. But far more often than we would like to admit, we choose what is in front of us, what is easy to reach, what the environment has made salient. The figs in the glass jar and the M&Ms in the opaque tub offer the same person the same options. What changed was not the menu and not the motivation. What changed was which option the room put in reach.
This is why placement beats persuasion so reliably. A poster is aimed at the deliberate, reflective part of us, the part that is often asleep at the moment of choice. Placement reaches the automatic, instinctive part, the part that is actually making the decision. Design for the reaching hand, not the reflecting mind, and the behaviour follows.
A word of honesty is owed here, because placement is not magic and it is easy to oversell. The research on choice architecture in shops and canteens is genuinely mixed. Some carefully run interventions move behaviour a great deal; others, equally careful, barely move it at all. A well-designed twelve-week trial that gave healthier cereals prominent shelf placement found no increase in sales of the healthier products, which is a useful reminder that visibility alone does not guarantee a shift. What seems to separate the interventions that work from the ones that do not is whether they change the genuinely automatic, low-deliberation choices, the handful of M&Ms grabbed in passing, rather than the considered, planned ones. Google's experiment worked because snacking is about as automatic as behaviour gets. The lesson is not that placement always wins. It is that placement wins where the choice is automatic, and that is exactly where posters and information lose.
Learn to apply this yourself
What makes canteen nudge design effective is that it targets behaviour at the point of least deliberation, where placement and salience do the work that no campaign or poster can. In the Behavioural Design Fundamentals Course, you learn to use the Influence Framework and SWAC to map exactly where automatic choices happen in any environment and redesign those moments on purpose.
View the Behavioural Design Fundamentals CourseWhat you can design this week
You do not need a behavioural science team or a New York office to use this. Salience and placement are everywhere a person chooses, and they are usually doing their work by accident. The move is to do it on purpose.
Make the good option the visible one. Whatever behaviour you want more of, ask where it currently sits in the field of vision and the field of reach. If it is at the back, behind a click, below the fold, out of sight, you are quietly arguing against it. Bring it forward. Salience is not a slogan; it is a position.
Add a little friction to the option you want less of. Google did not ban the chocolate. They moved it into an opaque tub. A small increase in effort or invisibility is often enough to shift the automatic hand, without removing anyone's freedom to choose. Friction is a dial, not a wall.
Stop relying on information to do placement's job. Calorie labels, nutritional charts and earnest signage speak to the reflective mind, which is rarely the one in charge at the moment of choice. They are not useless, but they are weak compared to where the thing actually sits. If you are leaning on information to change behaviour, you are probably under-using placement.
Check whether your room is arguing with your goal. This is the deeper audit. Every space, menu, interface and shelf is already a piece of choice architecture, whether or not anyone designed it on purpose. Walk through yours and ask what it is currently making easy. Very often you will find a wellness goal on the wall and a layout on the floor that quietly undoes it.
The thread is the same one that runs through all of this work, and through everything we do at SUE. The behaviour you want is rarely waiting behind a better argument. It is waiting behind a better arrangement. Google did not persuade anyone to eat fewer M&Ms. It moved the jars, and let the reaching hand do the rest. And the move cost almost nothing, which is the part that should make every leader pause. The wellness programme has a budget, a launch, a dashboard. The jar has none of that, and it did the work the programme was meant to do.
If you want to learn how to read a room the way Google read its canteen, and redesign the places where your people and your customers make their automatic choices, that is precisely what our Behavioural Design training is built to teach.
Frequently asked questions
Why did moving the M&Ms into an opaque container work when a health campaign would not?
Because snacking is an automatic behaviour, not a deliberate one. Nobody consciously decides to grab a handful of M&Ms; the hand reaches because the candy is visible, accessible and already on the way to the coffee machine. A health campaign addresses the reflective part of the brain, which is rarely in charge at that moment. Placing the chocolate in an opaque tub reduced its salience at the exact point where the automatic choice was happening, which is the only moment that actually counts.
Does choice architecture always work, or are there cases where it fails?
The article is honest about this: the research is genuinely mixed. A carefully run twelve-week trial giving healthier cereals prominent shelf placement found no increase in sales of those products at all. What separates the interventions that work from those that do not seems to be whether the target behaviour is truly automatic and low-deliberation. Google's experiment worked because grabbing a sweet while waiting for coffee is about as unconsidered as behaviour gets. Where people are making planned, considered choices, placement has far less leverage.
How do I audit whether my own office or space is working against my goals?
The article suggests a straightforward walk-through: look at what the layout is currently making easy, and ask whether that matches the behaviour you want to encourage. Very often you will find a wellness goal on the wall and a physical arrangement on the floor that quietly undoes it. The question to ask of any shelf, menu, interface or room is not "what does it say?" but "what does it make the reaching hand want to do next?"
What is the difference between adding friction and simply removing choice?
Google never banned the M&Ms. They moved them into an opaque tub, which introduced a small increase in effort without eliminating the option. That distinction matters: friction is a dial that shifts the automatic hand without coercing anyone, while removal takes the option away entirely. The effect on behaviour can be similar, but the ethical and practical implications are not. Friction respects freedom of choice while still doing most of what a ban would accomplish in low-deliberation situations.
If you want to learn how to design for behaviour rather than trying to motivate it, that's exactly what the Behavioural Design Fundamentals Course covers. View upcoming dates →
1.5 minutes on influence
Join 6,500+ readers · Free · Unsubscribe anytime